State of charge estimation of Li-ion batteries based on adaptive extended kalman filter

The extended Kalman filter (EKF) is widely adopted for the state-of-charge (SOC) estimation of batteries. The trial and error selection of noise covariance and variation of operating temperatures lead to convergence uncertainty and poor robustness of the EKF. This paper presents an adaptive EKF (AEKF) for online SOC estimation of lithium-ion batteries based on the Thevenin equivalent circuit model (ECM) that can mitigate the problems with EKF. The parameters of the first-order Thevenin ECM are estimated using the recursive least square (RLS) method at different operating temperatures. A pulse discharge test with lithium-iron-phosphate cell has been carried out in the LabVIEW platform, where SOC of the cell is determined by the coulomb counting method (CCM). Then the SOC is estimated using the EKF and AEKF methods and compared with the CCM method. The simulation and experimental results confirm that the AEKF shows better performance compared to the conventional EKF method.